In TrueSight Capacity Optimization (TSCO), what is the best way to import Business Driver data from custom 3rd party data sources for use in analysis and reporting? We have a 3rd party data source that provides a transaction rate per hour for an application. What is the best way to import that transaction rate (or other business driver data) into TSCO for use in reporting and analysis? For example, we'd like to create a Performance versus Load Analysis to see how well CPU utilization is corrected to the application transaction rate. |
The import of business driver data into TrueSight Capacity Optimization (TSCO) is generally going to be done using the Generic SQL Extractor or Generic CSV Parser ETLs. The product documentation describes the expected data format for each of these ETLs: Generic CSV parser: https://docs.bmc.com/docs/display/btco2002/Generic+-+CSV+file+parser Generic Columnar CSV parser: https://docs.bmc.com/docs/display/btco2002/Generic+-+CSV+columnar+file+parser Generic Database Extractor: https://docs.bmc.com/docs/pages/viewpage.action?pageId=914178206 The difference between the 'CSV parser' and the 'Columnar CSV parser' is that the CSV parser will import one metric name and value pair per line where as the Columnar CSV parser will import multiple metrics per line (with each metric in its own column). If you have a lot of different metrics to import for the same system for each interval the "Columnar CSV parser' is often the easier to implement. If you are importing a smaller number of metrics (particularly just one) or if you are importing a varied list of data for different intervals the 'CSV parser' is probably the easier to implement. Another feature to consider for data import is the "OpenETL Template" functionality which is described here: https://docs.bmc.com/docs/display/btco2002/Managing+an+Open+ETL+template That functionality basically adds a UI to assist with the creation of your CSV Parser or SQL Extractor ETL. Once an OpenETL template has been created you then create a CSV Parser or SQL Extractor based upon that OpenETL template to actually import the data. The following KA also provides some additional information and suggestions regarding the creation of a CSV Parser ETL: 000030572: TrueSight Capacity Optimization(TSCO) - Where can I find information on how to create a CSV file parser ETL? (https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=000030572) Looking at the product documentation for the CSV parser, the following example is provided for the input file format to import a business driver: TS;DURATION;WKLDNM;OBJNM;SUBOBJNM;VALUE;DS_WKLDNM 2009-01-31 01:00:00;3600;portal1;WEB_HIT_RATE;GLOBAL;126;portal1.bmc.com 2009-01-31 02:00:00;3600;portal1;WEB_HIT_RATE;GLOBAL;294;portal1.bmc.com 2009-01-31 03:00:00;3600;portal1;WEB_HIT_RATE;GLOBAL;292;portal1.bmc.com 2009-01-31 04:00:00;3600;portal1;WEB_HIT_RATE;GLOBAL;176;portal1.bmc.com In this example: TS: The Timestamp of the data point. This is the beginning of the interval this data represents DURATION: The Duration of the data point. This is the length of the interval this data point is associated with. WKLDNM: The name of the Business Driver entity this data point should be associated with. In this example the 'portal1' business driver entity will be used for this data. OBJNM: The name of the business driver metric SUBOBJNM: Generally for 'BY_*' metrics the name of the sub-object for this metric. For example, if the WEB_HIT_RATE were being reported separately for different applications each separate application may be named as a SUBOBJNM report the WEB_HIT_RATE metric separately for each. VALUE: The value reported for this business driver for this interval. DS_WKLDNM: For a "single lookup" business driver entity this will just be the name of the business driver. For a "multiple lookup" entity this can be a more complex representation of the business drivers name to ensure it is properly mapped. Generally for business drivers this will be a 'single' lookup. [See: 000215356: How Single lookup vs. Multiple lookup work in TrueSight Capacity Optimization (TSCO) ETLs (https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=000215356) for additional details -- although entity lookups is a relatively complex topic that often isn't an issue for business driver import] So, in general, you'll want to create a CSV Parser (or if the data is available in a 3rd party database a Generic SQL extractor) to obtain the data from your 3rd party data source. It is best for the input file format for the CSV Parser ETL to be designed to be specifically compatible with the Generic CSV parser ETL (even if that requires a custom script to massage the data that comes from your 3rd party data source into a better format before passing it to the Generic CSV parser ETL for import). |